Multiple secondary users can cooperate to increase the reliability of spectrum sensing in cognitive radio (CR) networks. However, the total transmission power grows approximately linearly with the number of cooperative secondary users. This study proposes a new approach to optimise the trade-off between sensing reliability and power efficiency in cooperative CR networks over fading channels. The authors assume K cooperative secondary users each collect N samples during the sensing time. The proposed approach is based on dividing the spectrum sensing into two phases. In the first phase, the authors use only n1 of N samples (n1 ≤ N), to check the channels state, then k of K cooperative secondary users (k ≤ K), which are in deeply faded channels are discarded. The authors call this n1 a check point of the sensing time. The spectrum sensing with relatively less-faded channels are continued during the second phase. Therefore there is a check point at which the sensing time can be optimised in order to maximise the probability of detection and the power efficiency. Several experiments are carried out to test the performance of the proposed approach in terms of probability of detection and power efficiency. The obtained results show that the proposed approach enhances the probability of detection and shortens the optimal sensing time. Moreover, it improves the overall power efficiency.